It's not often I come here with a serious puzzle, but...

Set the fuel-rev limit to 8500 and the fuel-cutting to begin at 8300 rpm.

Three runs, three successes, with the engine pulling freely to 8000 - that never happened before touching the rev limits. I didn't bother going higher because that wasn't the goal.

It's a little early to claim success since the "thing" only happens (or happened) intermittently. If maybe 10 runs can be performed without fail, then I'll consider the issue closed. As for whether or not to post on the ECU forum about this, not sure it'll matters. Because it is/was intermittent, Mr. Attitude Tech Guy can stick to his claim of perfection. If someone else posts that they're having the same problem, maybe I'll post, maybe I won't, because the tech guys have a bad habit of editing or deleting posts they feel aren't in their best interest. A case of sour grapes? Maybe a little.

On the way home from this morning's car thing, a new Mustang came barreling up behind me... ("oh I see how it's going to be..."). When he pulled over to pass, the 16-yr old in me couldn't resist "showing him." At the next light he shouted over, "your car is bad-ass!" It bothers me that as practically a senior citizen, such a childish statement still makes me feel good. Good thing the wife wasn't in the car...

You have anything but 0 in the rev limit pid values?

In reply to kb58:

You say you are recording each test- it should be really easy to show why it's happening, data always trumps opinion.

Paul_VR6 wrote: You have anything but 0 in the rev limit pid values?

I left the default values because I don't want the engine running without some kind of rev-limit. Even for this test I still want a safety net.

kb58 wrote:

Paul_VR6 wrote: You have anything but 0 in the rev limit pid values?

I left the default values because I don't want the engine running without some kind of rev-limit. Even for this test I still want a safety net.

Here's something else you can do- now that you know there's an effect- set the limits low to calibrate the control of the rev limiter. BTW, start with the I set low, and the D set to 0, and let the P do most of the work. D is rarely used.

So you can get that done pretty well, and set the final values where you want them.

alfadriver wrote: In reply to kb58: You say you are recording each test- it should be really easy to show why it's happening, data always trumps opinion.

Agreed, but unfortunately that is the problem. Nothing I have logged has directly shown the cause. The only factor that has had any affect on the problem has been changing the rev-limit values. (Changing the wastegate, boost control valve, and boost level didn't have any effect.) To be fair, there is still a chance that I'm incorrectly attributing "success" to the rev-limiter changes, when in fact it's something different. However, as time passes and more successful runs are made (with no failures) it's pointing more and more towards a bug in the ECU over-rev code.

I would think 0ing the pid would just disable the loop and make it a "hard cut" at the setpoint. Then again, l have no clue how this is supposed to work.

Fwiw in the manual there is a P value of 1 for fuel cut and the rest are 0'd.

That's what users might expect - I would - yet the ECU manufacturer said:

PID becomes active at rev limit minus window. If window is zero, PID becomes active at rev limit. If the PID are all zero, you'll have no rev limiter.

When I said that's a rather critical thing to not point out in the manual, I was informed:

If someone is going to do rev limit PID testing, I'd hope they'd be smart enough to do it at a lower rpm than their actual redline. Notice that we didn't put any warnings in the VE or ignition map section of the instructions because you could just as easily put bad values in those maps and damage an engine as well. These are stand alone engine management systems - we can't control what the end user decides to do.From the front page of the user manual: WARNING: This installation is not for the tuning novice! Use this system with EXTREME caution!

There's always a bit of an arrogant, belittling tone in this guy's responses. Unfortunately he seems to be the one who responds to my posts. Lucky me...

I'll post up a couple screenshots of yesterday's run in a bit; I know you guys love pictures. I'll also post a link to the datalogs.

A link to the full logs, though you'll need the AEMData logger app to view it.

http://www.midlana.com/Diaries/Current/ECU_00F.itlog

Screenshot from the above log, showing the first full-power run, which succeeds. The vertical blue line is 8000 rpm. Direct link to larger bitmap is http://www.midlana.com/Diaries/Current/8_28%20log.bmp

Screenshot from the above log, showing the second full-power run, which fails at 7700 rpm. Nothing changed between the first and second run and in fact took place roughly 5 minutes after the first. The engine had not been shut off and it was during the same drive. Direct link to larger bitmap is http://www.midlana.com/Diaries/Current/8_28B%20log.bmp

Anyhow, I think I've dragged this thread on long enough. I'll report back again after I'd done 10 runs. Thanks again for all the suggestions - it helps.

I love it, not for the novice or expert because we don't explain the parameters well!!

Good luck, hope this is it.

kb58 wrote: There's always a bit of an arrogant, belittling tone in this guy's responses. Unfortunately he seems to be the one who responds to my posts. Lucky me...

Seems to be par for the course for this sort of thing. I've encountered it with tech support for numerous other products. I had one chip-burner guy tell us that we had to change a customer's entire turbo and manifold setup, subtext "I don't like external wastegates, quit liking what I don't like"

You learn to roll with it. You either roll with it, or turn into an arrogant ass yourself. My back hurts too much to carry a big chip on my shoulder, so any time I have to provide tech support, I think of all the times I got bullE36 M3 like that and how much it sucks and then avoid doing that.

Thinking it over, I'm skeptical of a rev-limiter bug, but will take success where I can find it. I'll try and do a bunch of runs tomorrow to finish this up once and for all. (If you don't hear from me, contact the local jail...) If the runs are successful, I'm going to move on and try to not think about it too hard.

If it turns out to be that raising the rev-limiter has permanently fixed it, it appears that HiTempguy is the winner, as he was the first to cast an accusing glance at the rev-limiter. I realized that I had better go back and find who suggested it first, before the rest of you realize that you can edit your posts at anytime - and the forum here doesn't date-stamp your changes!

HiTempguy said: Aug. 25, 2015 2:57 p.m. ...Does the car kill spark or fuel at 8krpm? Maybe try raising the limiter a bit... just a thought. I seriously doubt that is the case though.

The good news: did 12 successful runs so I'm going to consider the issue closed.

The less-good news: Currently waiting on the side of the freeway for a tow. Had an oil system failure it was pretty exciting, smoke and oil everywhere, even on the inside of the windscreen. I suspect the dry sump saved the engine due to the extra volume and large pump section.

kb58 wrote: The good news: did 12 successful runs so I'm going to consider the issue closed. The less-good news: I think the O-ring in the remote oil filter adaptor blew out. Waiting on the side of the freeway now for a flatbed tow. It was pretty exciting, smoke and oil everywhere, even on the inside of the windscreen. I suspect the dry sump saved the engine due to the extra volume and large pump section.

Unfortunately, it looks like you're going to have to do a lot more 8000rpm high-gear pulls to verify the integrity of the oil system. Yeah. That's the reason.

I finally looked down to see at what speed redline happens in 3rd gear. Hmm, let's just say that if I'd been caught, a blown oil seal would be the least of my worries. I'm very aware that this sort of testing should be done on a dyno or at the track, but, well. As said before, such freeway runs are done with no passenger, no passing other cars at speed, and no fast lane changes. It's about as "safe" as it can be, though I know that doesn't mean anything if pulled over. So all in all, things are okay.

Pretty sure the large O-ring on the Canton remote filter adaptor blew out after the large hold-down nut backed off. The nut will be modified to be retained by safety wire.

Glad its fixed, but man what a way to park it!